The purpose of this chapter is to elaborate on the scientific methods used to collect data and information for this study. Methodology is the scientific methods used in a research project to collect data (Dalland 2012: 111). Structured methodology is a necessary component of the research process and research is done with the help of study, observation, experiment, analysis and comparison (ibid.:). This thesis draws on theoretical literature, an expert interview, media sources and five case studies to answer the research question and sub-questions presented in chapter one.
4.1 Qualitative research approach
There are two main approaches to research in social sciences; qualitative and quantitative methodology. Qualitative research is an empirical approach where research data can be categorized and classified, while quantitative data is measured in numbers and statistics (Bryman 2016) This research is a qualitative research study. Qualitative research studies aim to give an analytical description of a given situation; the aim might be to give a theoretical generalization.
Quantitative research studies aim to give statistical generalizations to describe a larger population, which can be measured in numbers (Bryman 2016: 149). Given the very nature of the research topic, and the research question being of exploratory character, the qualitative approach was deemed most beneficial for this kind of study. The research has been conducted as a combined desk-study, based on reports, news articles, academic books and journals, and qualitative interviewing. Qualitative studies are characterized by closeness and sensitivity due to the personal involvement of the researcher in the environment. Quantitative studies are characterized by distance and selectivity. Another aspect of qualitative research is the interpretation of findings. Quantitative interpretations often take on a more structured and precise form, while qualitative researchers extract key findings and make relevant interpretations (Dalland 2011: 115).
4.2 Research design
The research design provides a framework for the collection and analysis of data and refers to the strategy a researcher use for the study. Choosing what type of design depends on the subject and research question. A case study is an in-depth, contextual examination of a specific phenomenon (Yin 2003). The purpose of a case study is not to find a universal explanation but rather clarify the realities of a specific case. A case study investigates a person, place, event, period, phenomenon or other type of subject to extract key themes (ibid.). This research investigates the phenomenon of female jihadists in Europe. That in itself is a single and unique phenomenon.
Still, in order to compare how the roles of these women have changed and for what reasons over the last two decades, a selection of representative cases from each era have been picked out for analysis. Each of these cases are also individual case studies, and each decade could also be applied as a single case study. Therefore, each of the individual jihadists presented in the analysis is a ‘case’ and I am comparing five cases in order to answer the research questions.
Hence, a multiple case study design is applied. A multiple case study design allows the researcher to study differences and similarities between a selected sample of cases within a research area, or topic (ibid.: 45). This approach has distinct advantages and disadvantages. One advantage is that the evidence, or findings, from multiple cases are often considered more compelling thus making the overall study more robust. Also, it allows the researcher to analyze the data both within each case and across cases, or situations. The researcher can thus make extensive comparisons and yield more findings that shed light on the overall topic. A disadvantage is that this approach is often time-consuming, expensive and requires extensive amount of data and resources (ibid.). Still, evidence founded in a multiple case study is often considered stronger and more reliable compared to a single-case study. However, this is naturally dependent on the quality of the data and analysis made by the researcher.
4.3 Multiple case study design
The female jihadists in Europe is the research topic of this study. Over the last two decades, female participation in jihadist activity has increased taneouslytounsly as the overall terrorist activity in Western Europe has increased. As jihadist terrorism has historically been preserved for men, it is both an interesting yet worrisome development which up until now, has received limited attention. Therefore, this study aim to be a contribution to the understanding of female terrorism, and of terrorism in general. By studying gender structures and gender ideology of
jihadist ideology, representative cases from the last two decades whilst applying a typology created by one of the leading experts on jihadist terrorism in Europe, the research aims to contribute to the understanding of female jihadism and provide insight into how jihadist women in Europe is a growing phenomenon and why our understanding of terrorism should include a gender perspective. The design of this research, as described above, is comprised of several cases of female jihadism in Europe from the last two decades. I decided on three criteria for choosing which cases to include in the analysis. First, the women in each case had to have been sentenced, and found guilty, in a criminal court. Additionally, the cases had to be related to jihadist terrorism, preferably cases where the women had been sentenced on terrorist-related charges.
One of the cases is an exception, where the subject was charged with attempted murder. The woman was nevertheless inspired by jihadist ideology to commit the crime, and therefore, this research considers her a female jihadist. Second, the cases had to be European, meaning the women had to be based in Europe, the crime was planned and/or committed in Europe and the women were sentenced by a European court. Third, the cases had to have occurred between the year of 2000 and 2018.
4.4 Data collection methods
In order to conduct a study, the researcher needs to collect information on his or hers research area, so-called data. According to Bryman (2012), there are different types of data and different ways of collecting them. Primary data is data collected by the researcher, for instance from interviews, whilst secondary data is already existing data which has been collected and interpreted by someone else in the past. For this study, the data collection started after reviewing the literature which led to the creation of the research question. This study relies on both secondary and primary data, which combined result into triangulation and refers to using more than one method of data in the study of a social phenomena. Triangulation can help identify strengths and weaknesses of each method in researching a phenomena, and if two methods indicate the same results, it strengthens the research’s credibility (ibid.: 383-386).
Secondary data has already been collected by other researchers and the data has been interpreted (Dalland 2012). An advantage of using secondary data is that it saves time and costs.
Additionally, it can be used to compare data over time (e.g. longitudinal studies). For this study, secondary data was not only beneficial but also completely necessary, due to the complexity of
the topic and limitations in terms of available informants. Due to security obstacles, the timeframe and lack of accessibility to actual individuals with first-hand knowledge, interviews with the female jihadists themselves was not a viable option for this study. These women are, not surprisingly, not easily accessible and they are often part of networks that can be considered quite dangerous. Therefore, this research relied heavily on former studies conducted on this topic, and on women connected to terrorist networks. Primary data is new data collected by researchers themselves. Primary data is any information that has not been interpreted or analysed by other researchers. In qualitative research, the dominant method of collecting primary data is through interviews. Yet, primary data can also be in the form of news articles (at the time of an event), autobiographies, original documents e.g. law documents or personal journals, letters and audio and video footage. The primary data of this study consist of one expert interview, published interviews with female jihadists (conducted by others), audio recordings and news sources.
4.5 Expert interview
I conducted a telephone interview with one of Europe’s leading experts on jihadist terrorism in Europe, Petter Nesser, who also created the typology which was applied as a framework in the analysis of the individual women in the case studies. Telephone interviews have been relatively common in quantitative research but not so common in qualitative research (Bryman 2012: 484).
There are numerous challenges to keep in mind when using this method of interviewing. First, the researcher does not get to witness and observe the setting in which the interviews takes place, and therefore not the body language of the interviewee. Second, technical obstacles could arise, such as bad connection or troubles recording the interview. Third, it is easier for the interviewee to abruptly end the interview if she or he wants to. Still, there are nevertheless benefits of doing telephone interviews. It is cost-effective, and cheaper than compared to face-to-face interviews and it is time-effective. It can also be effective in the sense that the interviewee may feel less stressed and anxious when the interviewer is not physically present, and still deliver comprehensive and expansive replies (ibid.:). In this case, a telephone interview felt pertinent considering this was an interview with an expert whose research I was already quite familiar with, in addition to it being a more formal, or academic, type of interview whereas I believed the observations of the surroundings or his physical responses to my questions would not be crucial for the analysis of the interview.
There are two main types of qualitative research interviews; unstructured and semi-structured interviews (Bryman 2012: 466). Unstructured interviews are quite similar to a conversation, where the researcher asks a question and follows up based on the reply of the respondent with questions that seem relevant. Semi-structured interviews use an interview guide prepared beforehand by the researcher containing questions and topics that should be covered. These questions can be altered and their order adjusted according to the flow and progress of the interview (ibid.: 467). I used a semi-structured interview with a interview guide prepared beforehand. Considering this was an interview focusing on a complex topic with a person whose in-depth knowledge and expertise relating to the research area is extensive, it was important that I as a researcher was prepared and had created questions relevant to his expertise yet also to my research question. A semi-structured interview guide helps the researcher to stay on topic whilst still open up for follow-up questions and more flexibility depending on the replies. The interview took place on the 10th of April in the morning. I recorded the interview with the permission of the interviewee. In the beginning, I asked Nesser questions about the overall situation of jihadism in Europe, and about the different networks and cells operating across the region. Following, he gave me a relatively detailed account of how women had been involved in Europe, from the early 1990’s to date. After a while, I directed the questions more towards his typology and how, and if, he believed it could be applied in the study of female jihadists, considering it was created based on data collected on male jihadists. In the end, we talked about the overall threat of jihadism in Europe, now and for the future. My impression was that the interview went well and fluently, Nesser gave extensive and in-depth answers to my questions and the interview lasted for nearly an hour.
4.6 Typology of female jihadists in Europe
A problem that arises when applying a typology that was created and based on data depicting male jihadists, is that women do not necessarily meet the standards set out by studies of male terrorists (Jacques and Taylor 2013). It also risks falling into the trend of applying gender as a way of describing differences in sexes instead of as an analytical tool in describing relationships of power and structures. On the other hand, considering studies on male and female terrorists have shown similarities and common characteristics between the genders, using an already existing typology can be beneficial. It not only provides the researcher with a framework that is
already based on extensive data collection from years of research but it also provides an etiology of male terrorists that can be used as a comparison to the female cases to illustrate the relation of gender and power, and the differences in the etiology of male and female terrorists.
Simultaneously, it provides a background of how those structures of power in jihadist movements has been, and still is, dominated by men thus making it easier to detect changes or compromises in the gender structures of those movements. Still, considering the analysis of this study is to a great extent based on media representations of these women, the descriptions cannot be taken at face value. The mainstream media has, as pointed out by several scholars, a tendency to sensationalize cases in a way that does not include scientific methods of analysis. These representations must therefore be critically analysed and coded, before being applied with an analysis based on a theoretical framework.
4.7 Analysis of qualitative data material
The data collected for the first unit of analysis in this study consisted mainly of written texts in the form of police reports, academic reports, news coverage and articles of the cases in question, as well as other secondary sources. The second unit of analysis consisted mainly of written texts in the form of interview transcripts, both the one I conducted myself and previous interviews with some of the female jihadists, including police interview transcripts. Unlike the statistical analysis of quantitative data, there are no clear-cut rules for the analysis of qualitative data (Bryman 2012: 570). Qualitative data analysis is more flexible; the researcher usually analyses the data in the same time period as the data collection takes place: impressions, observations, and thoughts from for instance the interviews are usually more vivid and memorable right after the interviews have taken place. Qualitative data material, such as interview transcripts, can amount to a vast number of pages with written words. The purpose of analysis is to uncover patterns in the data. Coding is a common method used to find these patterns (Saldana 2015). A code is a keyword that can describe or characterize a meaning or theme in a given sentence or paragraph in a text. Codes can be descriptive or interpretive: the latter will be subjective as it is the researcher’s understanding of the text, while descriptive codes describe the actual content of the text (ibid.: 3).
I used coding in order to determine content meaning and themes. After the expert interview, I transcribed the interview into a text and annotated it, marking relevant information, reoccurring
themes, and overall meanings in the paragraphs. Additionally, I compared the results to the interview guide to assess if I had managed to collect answers to all of the questions. Of course, I had not. In semi-structured interviews, the flow and direction of the conversation will not be known beforehand and the researcher can try to redirect the conversation back to the topic.
However, this does not mean that the unintentional change in direction is negative: the interview subject may have additional information or beneficial insight for the study that the researcher did not expect. In regards to the interviews by other researchers with the women of the case studies, I followed the same procedure. I coded the women’s statements and replies, trying to find recurring themes and patterns.
4.8 Quality criteria
Regardless of what type of research design, or methodology, is used in a study, it is necessary to apply some common evaluation criteria in order to measure the credibility and quality of the research (Dalland 2012: 120). In quantitative research, validity and reliability are two essential quality control criteria. The first refers refers to whether “a measure of a concept really measures the concept”, meaning if the research methods are credible and the second refers to which degree a research is replicable, meaning that another researcher should be able to conduct the same study with the same methods and yield somewhat the same results (Bryman 2012: 156-158). The quality of the data collected must be seen in relation to the purpose of the study, which is to shed light on the research question. For data to be valid, it must be relevant for the study. This concerns both the secondary data used such as academic literature and primary data such as the interviewees who participate in the research. Reliability refers to the trustworthiness of the research. It means that repeated measures with the same instrument on a given sample of data should yield similar results, meaning the reliability of a study will be high if one gets similar data when repeatedly collecting data concerning the same phenomenon (ibid.:). This is difficult in qualitative analysis, because the researcher is subjective and interpretive in his or hers analysis.
Therefore, it is impossible for another researcher to yield the exact same results.
Due to these difficulties, an alternative method of evaluating the quality of qualitative research was introduced by Guba and Lincoln in the 1980’s (Morse 2015). Their method introduced an alternative terminology for assessing qualitative research to ensure rigor; trustworthiness and authenticity. In simple terms, they argued that there can be more than one ‘truth’ in the study of
the social world, and therefore rejecting the presupposed idea of one ‘absolute truth’ replicable by the standard criteria of reliability and validity as prevalent in quantitative research (ibid.:).
Nevertheless, while the actual analysis of a topic in qualitative studies will differ in terms of analytical methods, it should be possible for another scientist to replicate the same findings.
Qualitative researchers do not use the same instruments to measure validity and reliability so therefore, they introduced the method to ensure that the research is credible, transferable, confirmable, and dependable. If these four criteria are established in a qualitative study, the overall trusthworthiness of the research is ensured (ibid.:).
4.9 Limitations of the data collection
The main limitation is the fact that the persons in question being studied, the female jihadists, are being studied from the outside based primarily on data from other sources than the actual women themselves. As mentioned, interviews with these women was not an option for this study which limits the contribution of this research in terms of understanding these women from their own perspective. I chose to focus on collecting data from both secondary and primary sources already available instead of conducting my own interviews with scholars, experts or people with knowledge of this topic. I decided on this approach for two reasons. First, I believed that if I could not hear it from the individuals in focus themselves, the female jihadists, it would be beneficial to dedicate that time to locate sources whom had already done this, for instance
The main limitation is the fact that the persons in question being studied, the female jihadists, are being studied from the outside based primarily on data from other sources than the actual women themselves. As mentioned, interviews with these women was not an option for this study which limits the contribution of this research in terms of understanding these women from their own perspective. I chose to focus on collecting data from both secondary and primary sources already available instead of conducting my own interviews with scholars, experts or people with knowledge of this topic. I decided on this approach for two reasons. First, I believed that if I could not hear it from the individuals in focus themselves, the female jihadists, it would be beneficial to dedicate that time to locate sources whom had already done this, for instance